PLESS: Pseudo-Label Enhancement with Spreading Scribbles for Weakly Supervised Segmentation

TL;DR

PLESS enhances pseudo-label quality in weakly supervised segmentation by propagating scribble information within spatially coherent regions, leading to improved accuracy across multiple algorithms and datasets.

Contribution

It introduces a hierarchical region-based pseudo-label refinement method that is model-agnostic and improves segmentation performance in scribble-supervised learning.

Findings

Consistent accuracy improvements on cardiac MRI datasets

Effective integration with various pseudo-label algorithms

Code will be publicly available on GitHub

Abstract

Weakly supervised learning with scribble annotations uses sparse user-drawn strokes to indicate segmentation labels on a small subset of pixels. This annotation reduces the cost of dense pixel-wise labeling, but suffers inherently from noisy and incomplete supervision. Recent scribble-based approaches in medical image segmentation address this limitation using pseudo-label-based training; however, the quality of the pseudo-labels remains a key performance limit. We propose PLESS, a generic pseudo-label enhancement strategy which improves reliability and spatial consistency. It builds on a hierarchical partitioning of the image into a hierarchy of spatially coherent regions. PLESS propagates scribble information to refine pseudo-labels within semantically coherent regions. The framework is model-agnostic and easily integrates into existing pseudo-label methods. Experiments on two public cardiac MRI datasets (ACDC and MSCMRseg) across four scribble-supervised algorithms show consistent improvements in segmentation accuracy. Code will be made available on GitHub upon acceptance.